SuperAPI AI Directory
Partner with us中文AI newsFavorites
Skillsxiaoqi-fitness-tracker

xiaoqi-fitness-tracker

Office & productivity

View repository

Updated 10-15 09:39

About this skill

Fitness Tracking (General Edition)

A complete system to help users manage fitness data, track body changes, log diet, and generate training plans.

First use: profile creation guide

Check whether 05_健身/_身体档案.mdexists in the working directory. If it does not exist, enter the profile creation process.

Profile creation questions (ask one at a time, do not throw them all out at once)

  1. Basic information: height, current weight, age, sex
  2. Current training status: How many times per week do you train? Do you have a coach? Do you train on your own or follow classes? Roughly how long each time?
  3. Training goals(multiple choices allowed):
    • A. Fat loss
    • B. Muscle gain/body shaping
    • C. Posture improvement
    • D. Physical performance improvement
    • E. Emotional/mental health improvement
    • F. Other (user-described)
  4. Body composition data: Do you have a body composition report (Visbody, InBody, etc.)? If so, send a screenshot or file. If not, use the basic information to create the profile
  5. Diet situation: Do you have any special dietary needs or restrictions? (Vegetarian, allergies, medications affecting appetite, etc.)
  6. Protein target: Automatically calculate the recommended value based on weight and goals; user confirms or adjusts
    • Muscle gain recommendation: 1.6-2.0g/kg body weight
    • Maintenance recommendation: 1.2-1.4g/kg body weight
    • Fat loss recommendation: 1.4-1.8g/kg body weight

After profile creation is complete

Automatically create the following file structure:

05_健身/
├── _身体档案.md    ← 基线数据 + 体测记录 + 每日体重 + 阶段分析
├── _训练计划.md    ← 阶段目标 + 周训练框架 + 训练原则
├── _训练日志.md    ← 每次训练记录
├── _饮食日志.md    ← 每日饮食 + 营养估算
└── 看板.html       ← 可视化仪表盘

The dashboard is generated using the design system of the xiaoqi-board skill (light and dark dual themes, responsive). Sections include:

  1. Core metric cards
  2. Daily weight check-in (7-day cards)
  3. Today's diet (protein progress + macros + meal details)
  4. Phase goal progress bar
  5. Weekly body composition comparison (tab card switching)
  6. Body composition detail table (if body composition data is available)
  7. Segmental distribution (if available)
  8. Circumference data (if available)
  9. This week's training overview
  10. Background and strategy

Sections without body composition data are not generated, and are automatically added after follow-up measurements.

Six major functions

1. Body composition data entry

Trigger: The user sends body composition screenshots or report files (Visbody, InBody, body fat scale app screenshots, etc.)

Process：

  1. Identify the data source and extract all available data
  2. Append to _身体档案.md, creating a new entry by date
  3. Compare with the previous data, calculate the amount and direction of change
  4. Update phase analysis
  5. Update the dashboard (core metrics + weekly comparison tabs + detail table + segments + circumference + goal progress)
  6. Output a brief analysis

Supported data sources：

  • Visbody: screenshot or HTML report
  • InBody: screenshot
  • Body fat scale app screenshot (Huawei, Xiaomi, Youpin, etc.)
  • Data dictated by the user

Comparison logic(direction judgment):

  • Skeletal muscle/muscle mass ↑ = good
  • Body fat percentage/body fat ↓ = good
  • Waist-to-hip ratio ↓ = good
  • Basal metabolic rate ↑ = good
  • Visceral fat level ↓ = good
  • Whether weight is good or bad depends on the user's goal (↓ is good during fat loss, ↑ may be good during muscle gain)

2. Daily weight check-in

Trigger: The user reports a weight number, or says "check in" "today XX kg"

Process：

  1. Update _身体档案.md the daily weight table for the current week
  2. Update the card for the corresponding date on the dashboard
  3. Respond in one sentence, without lengthy analysis

Note：

  • Use naked weight for weight (after waking and emptying). Remind the user to keep measurement conditions consistent
  • The weight in a body composition report may be clothed weight; a difference from naked weight is normal
  • Automatically switch to a new week every Monday

3. Diet record

Trigger: The user sends food photos, takeout screenshots, or dictates what they ate

Process：

  1. Identify food types and approximate portions
  2. Estimate protein (P), fat (F), carbohydrates (C), and calories (kcal) for each item
  3. If the user says they only ate part → scale proportionally
  4. Write to _饮食日志.md the corresponding date and meal
  5. Update the "Today's Diet" section of the dashboard
  6. Gently remind when protein is insufficient and give supplement suggestions

Meal: Breakfast / Lunch / Dinner / Snack. Judge based on time or the user's explanation.

Estimation principles：

  • Rather slightly overestimate protein to give the user positive feedback
  • The oil in takeout/restaurant food is usually 20-30% higher than home cooking
  • If the user says "ate half" → strictly follow the proportion
  • Mark uncertain parts as "estimated"
  • Common foods can be given directly; there is no need to explain the estimation process item by item

4. Training plan generation

Trigger: "I want to train on my own" "I'll train on my own tomorrow" "give me a plan" "what should I train"

Process：

  1. Read the most recent training records
  2. Read the current phase goals and training principles
  3. Read the latest body composition/weight data
  4. Generate a detailed plan for the specified duration

Plan format：

## 训练方案 | MM.DD 周X | [训练主题]

### 热身（5-8 min）
- 动作名 | 时长/次数 | 要点

### 主训练（主体时间）
#### 模块 A：[部位]
- 动作名 | X组 x X次 | 重量建议 | 组间休息 | 节奏提示
（每个动作附一句要点说明）

#### 模块 B：[部位]
...

### 收尾（5-8 min）
- 拉伸/放松动作

### 今日重点
一句话核心注意事项

Programming principles：

  • If there are coach sessions, make self-training and coach sessions complementary
  • Prioritize compound movements (squats, deadlifts, presses, rows)
  • Progressive overload: refer to the last training record and try to increase
  • Train each target muscle group at least 2 times per week
  • Set 1-2 achievable small goals (reps/weight PRs) to create immediate sense of achievement
  • Adjust intensity according to the user's current training level—be conservative for beginners, progress normally for those with a foundation

5. Coach sessions / class follow-along records

Trigger: The user says "today's coach session covered..." "just finished class" "followed a group class" + content description

Process：

  1. Extract training content: body part, exercise, intensity, duration
  2. Write to _训练日志.md
  3. Update _训练计划.md training log summary
  4. Update the corresponding training card on the dashboard
  5. Brief response

6. Dashboard Update

After each data change, update the changed parts in 看板.html at fixed points:

Data Change Update Location
Weight check-in Daily weight card
Diet record Today's diet section
Body measurement data Core metrics + weekly comparison + detailed table + segments + circumferences + goal progress
Training record This week's training card

Automatically update the week marker every Monday, and switch the daily weight section to the new week.

Interaction Style

  • Weight check-in, training record: one sentence, no verbosity
  • Diet record: list estimate + protein target reminder (when insufficient)
  • Body measurement analysis: brief comparison + trend judgment, 2-3 lines
  • Training plan: complete and detailed, ready to use for training
  • Daily conversation: do not bring up fitness proactively, unless the data shows a change that requires a reminder

Notes

  • Medication information is highly private; record it only when the user proactively tells you, and do not ask follow-up questions proactively
  • Training plans do not replace medical advice; when injuries or pain are involved, recommend consulting a professional
  • Weight fluctuations are normal; do not overinterpret changes from a single day
  • Respect the user's dietary choices; focus on recording, and do not make moral judgments